Tunable Tool-Call Rates in LLM Agents via Representation Steering

• URL: http://arxiv.org/abs/2608.25198v1
• Date: Tue, 25 Aug 2026 22:35:18 GMT
• Title: Tunable Tool-Call Rates in LLM Agents via Representation Steering
• Abstract summary: We show that a tool can be controlled by a single linear direction in its residual stream, extracted without any training from the model's own tool-use preference signal.<n>We also show that the direction generalizes to unseen tools with strength comparable to each tool's own direction without favoring any specific tool choice.
• Score: 45.332672398938136
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Deciding whether to call a tool is a core competence of an LLM agent, and a costly one to get wrong: needless calls add latency, accrue cost, and may trigger irreversible side effects, while missing calls leave the model confidently wrong on questions it could only answer through tool-calls. Models manage this balance poorly, both over-using and under-using tools. Existing methods such as post-training and prompt engineering are expensive and difficult to modify at inference time. We show that whether an instruction-tuned model calls a tool can be controlled by a single linear direction in its residual stream, extracted without any training from the model's own tool-use preference signal and turned into an inference-time intervention with no prompt change. Adding the direction with strength $α$ moves the call rate monotonically from near $0\% $ to over $90\%$ while keeping calls well-formed. The steering works in both directions: dialing it down suppresses calls, and dialing it up induces new calls that land precisely on the questions the model cannot answer from its own knowledge. We also show that the direction generalizes to unseen tools with strength comparable to each tool's own direction and without favoring any specific tool choice. With live tool execution, a single sweep of the steering traces a cost/accuracy Pareto frontier and nearly doubles open-domain QA accuracy ($0.29 \! \rightarrow \! 0.56$); the same recipe transfers across a diverse range of models spanning dense, MoE, and multimodal architectures, without any training. Our code is publicly available at https://github.com/YuqiChen4188/Steering-Tool-Use-Propensity.

Related papers

• Joint Optimization of Tool Creation and Use for Large Language Model Agents [69.07066297088187]
  SMITH (-grounded Multi-task Iterative Tool Honing) trains tool creation and tool use inside a single policy.<n>A 4B Qwen3 trained with SMITH on 13 procedural reasoning tasks with exact verifiers reaches 79.8 macro-average accuracy on held-out tasks.<n>It also reaches 40.4 on TabMWP-Hard and 42.6 on out-of-domain GQA (+7.6 over the best same-backbone inference-time baseline)
  arXiv  Detail & Related papers  (2026-08-25T13:59:34Z)
• Agentic Routing: The Harness-Native Data Flywheel [35.444792826526]
  Report studies singleton and multi-model routing on agentic benchmarks including DRACO and PinchBench.<n>It argues that agentic routing is not merely cost control, but a data engine for agent-native training.
  arXiv  Detail & Related papers  (2026-07-13T11:05:55Z)
• Diagnosing and Calibrating Tool-Call Boundary Drift in Multi-Teacher On-Policy Distillation [6.100373666222075]
  We show that multi-teacher on-policy distillation can induce a behavior shift invisible from aggregate losses alone.<n>In a two-teacher tool-use setting, vanilla generalized knowledge distillation improves tool-call recall but also moves the model toward over-calling.<n>We propose Soft Clamp, a per-token divergence calibration method that dynamically compresses extreme token-level Jensen-Shannon divergence.
  arXiv  Detail & Related papers  (2026-07-08T06:26:13Z)
• ToolGate: Token-Efficient Pre-Call Control for Tool-Augmented Vision-Language Agents [9.793491545168575]
  ToolGate predicts execute/skip decisions from trajectory text and simple structural features.<n>Across two Qwen3-VL backbones, ToolGate reduces token cost to 64-69% of the unrestricted ReAct baseline.
  arXiv  Detail & Related papers  (2026-06-02T02:44:27Z)
• Model-Adaptive Tool Necessity Reveals the Knowing-Doing Gap in LLM Tool Use [47.29360932085394]
  Large language models (LLMs) increasingly act as autonomous agents that must decide when to answer directly vs. when to invoke external tools.<n>We introduce a model-adaptive definition of tool-necessity, grounded in each model's empirical performance.<n>We find substantial mismatches of 26.5-54.0% and 30.8-41.8%, respectively.
  arXiv  Detail & Related papers  (2026-05-13T18:59:28Z)
• LLM Agents Already Know When to Call Tools -- Even Without Reasoning [25.40369702634587]
  LLM agents tend to call tools indiscriminately, even when the model can answer directly.<n>No existing benchmark systematically studies when a tool call is actually needed.<n>We propose When2Tool, a benchmark of 18 environments spanning three categories of tool necessity.
  arXiv  Detail & Related papers  (2026-05-10T01:37:40Z)
• ET-Agent: Incentivizing Effective Tool-Integrated Reasoning Agent via Behavior Calibration [68.89572566071575]
  ETAgent is a training framework for calibrating agent's tool-use behavior.<n>It is designed to progressively calibrate erroneous behavioral patterns to optimal behaviors.
  arXiv  Detail & Related papers  (2026-01-11T11:05:26Z)
• Learning to Route LLMs from Bandit Feedback: One Policy, Many Trade-offs [69.2486294522259]
  BaRP is a Bandit Routing-feedback with Preferences approach that trains under the same partial-feedback restriction as deployment.<n> Framed as a contextual bandit over prompt features and a user preference vector, our method simulates an online feedback setting during training and adapts its routing decisions to each new prompt.
  arXiv  Detail & Related papers  (2025-10-08T18:24:59Z)
• Efficient Tool Use with Chain-of-Abstraction Reasoning [63.08202389132155]
  Large language models (LLMs) need to ground their reasoning to real-world knowledge.<n>There remains challenges for fine-tuning LLM agents to invoke tools in multi-step reasoning problems.<n>We propose a new method for LLMs to better leverage tools in multi-step reasoning.
  arXiv  Detail & Related papers  (2024-01-30T21:53:30Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.